DESTRUCTION OF CONCRETE BY AEKALI 
7 
The knowledge of the fact that our drain waters contain carbonates led 
me to try the action of a mixture of sodic sulfate and carbonate on the 
coarsely ground concrete. This mixture attacked the cement rapidly yield¬ 
ing a white decomposition product rich in carbonic acid. This result is no 
other than what we would expect. If the tri-calcic silicate in setting really 
liberates four molecules of calcic hydrate, as we suppose the reaction should 
be accompanied by the liberation of a corresponding amount of sodic 
hydrate. The following formulae will make this point clear: 
2(3Ca0.Si0,)+9H,0==2(Ca0 Si0,)5H,0-j-4Ca(0H),. 
The calcic hydrate liberated or very loosely held can readily act on 
the sodic carbonate as indicated by the following simple formula; 
4 Ca( 0 H) 2 -f 4 Na 2 C 03 =: 4 CaC 03 + 8 Na 0 H. 
The sodic hydrate thus freed is capable of further attacking the silicates 
and possibly the aluminates, carrying silicic acid, alumina and lime into 
solution, probably as hydrated silicates. The treatment of the cement with 
a mixed solution of sodic carbonate and sulfate showed that it is a fact that 
calciic carbonate is formed and also that silicic, alumina and lime do 
go into solution and that too without the aid of free carbonis acid. 
No care was taken to make the solution of mixed salts even approxi¬ 
mately of the same strength, as the ground waters, and no ratio between the 
salts was observed, simply, that the solution was a weak one; two to three 
grams of sodic carbonate to the liter with five or six grams of the sulfate 
The time of the experiment was 14 hours. The time was short, and the 
solution not as rich in salts as some of our ground and drain waters. The 
tiles on the other hand are exposed to the action of similar solutions, some 
weaker but others stronger from the time of being laid till taken out, in 
this case eight to nine months. In view of these considerations it is not 
strange that the mass of concrete should be softened and destroyed. 
Using the results of these rather crude experiments we may interpret 
our analyses as indicating that the cement has been deprived of silicic acid; 
it originally contained 22.50 per cent, but the hydrated decomposition pro¬ 
duct contains only 10.98 per cent or 14.16 per cent calculated on a water 
free basis. While the ratio of the weight of the decomposition product to 
the original weight of the cement is not known, there is no'evidence that 
there has been any increase in the weight which would invalidate the pre¬ 
ceding observation and a second experiment with cement and a solution 
containing sodic sulfate and carbonate leaves no question about the removal 
of silicic acid, alumina and lime by waters containing these salts. The 
analyses also suggest the removal of a relatively large amount of lime. 
The cement contains 59.56 per cent, of lime while the alteration product 
contains but 37.58 per cent, (calculated on a water free basis). Further the 
ferric and aluminic oxids have been reduced to about one half as much in 
the decomposed cement as there was in the original. The sodic sulfate, 
more particularly the calcic and magnesic sulfates which are present in the 
ground water will at the same time act on the tri-calcic aluminate, forming 
the sulfo-aluminate 2(3CaO Al203)-|-5Ca0S03, which crystallizes with 60 
molecules of water and is soluble. That these reactions really have taken 
place is indicated by the increase of the sulfuric acid SO3 which amounts 
to 20.07 per cent, and likewise of the carbonic acid which amounts to 12.16 
